Munich Personal RePEc Archive
Well-being and intended early retirement among older European workers: does job satisfaction matter? A 6-Wave follow-up
Cantarero-Prieto, David and Pascual-Sáez, Marta and Blázquez-Fernández, Carla
Universidad de Cantabria
2018
Online at https://mpra.ub.uni-muenchen.de/89077/
MPRA Paper No. 89077, posted 24 Sep 2018 18:34 UTC
1
Well-being and intended early retirement among older European workers: does job satisfaction matter? A 6-Wave follow-up
David Cantarero-Prieto and Marta Pascual-Sáez and Carla Blázquez-Fernández Department of Economics, University of Cantabria. Avda. Los Castros, s/n, Santander CP 39005. Spain. E-mails: david.cantarero@unican.es; marta.pascual@unican.es;
carla.blazquez@unican.es
Corresponding author: E-mail: carla.blazquez@unican.es
Abstract: In recent years, population aging has received great attention in developed countries given the social challenges that it entails. At this regard, it is well documented that this collective is associated with fewer resources (both physical and economic).
Furthermore, ageing societies incite an increase in the inactive population and so, threaten the financial viability of the social protection systems. This study investigates the effects of different factors on early retirement intentions among European workers aged 50-65 using the latest available data (waves 1-6: 2004-2015) from the Survey of Health, Ageing and Retirement in Europe (SHARE). We shed new light on this causal relationship controlling for job characteristics and well-being indicators. Our empirical results based on logistics regressions suggest that people that is satisfied with their jobs (OR = 0.61; 95 % C.I. 0.53, 0.71), with very high appreciation of their quality of life (OR = 0.56; 95 % C.I. 0.49, 0.64) or with good health (OR = 0.55; 95 % C.I. 0.47, 0.65) would have less intentions of early retirement, that is, decreased odds of work exit. Besides, social-environment would matter.
Keywords: Early retirement intentions; Job satisfaction; Quality of life; Health;
SHARE; Panel.
JEL Classification: I10; J26; J28.
2
Disclosure of potential conflicts of interest
Ethics approval and consent to participate
Ethics approval is not required for this paper, since we did not collect data with personal information. The paper is the result of a research carried on independently by the authors. No plagiarism and no conflict of interest can be addressed to this research.
Availability of data and material
‘Not applicable’.
Competing interest
The authors declare that they have no competing interest.
Funding
‘Not applicable’.
Authors’ contribution
All authors contributed to the writing of the manuscript and read and approved the final manuscript.
Acknowledgements
This paper uses data from SHARE Waves 1-6 (http://www.share-project.org/data-documentation/Waves- overview/Wave-6.html), see Börsch-Supan et al. (2013) for methodological details. The SHARE data collection has been primarily funded by the European Commission through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3: RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE: CIT4-CT- 2006 028812) and FP7 (SHARE-PREP: N°211909, SHARE-LEAP: N°227822, SHARE M4: N°261982).
Additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, HHSN271201300071C) and from various national funding sources is gratefully acknowledged (see www.share-project.org).
This paper uses data from the generated easySHARE data set (DOI:10.6103/SHARE.easy.600), see Gruber et al. (2014) for methodological details. The easySHARE release 6.0.0 is based on SHARE Waves 1, 2, 3 (SHARELIFE), 4, 5 and 6 (DOIs: 10.6103/SHARE.w1.600, 10.6103/SHARE.w2.600, 10.6103/SHARE.w3.600, 10.6103/SHARE.w4.600, 10.6103/SHARE.w5.600, 10.6103/SHARE.w6.600).
3
1. Introduction
The increase over time in the proportion of older people above the total population has received considerable attention in recent years.1 In order to manage this demographic challenge, the main concern is related to the financial sustainability of our welfare states.2 At this regard, what it is well known is that the socioeconomic profile of the elderly people is determined by three main elements: consumption, income and free time.
Accordingly, that ageing population is associated with i) a greater intensity and frequency in the consumption of social-health services and long-term care, ii) pensions as main source of income and iii) greater availability of leisure time. Therefore, it can be established that the main issues for policy makers associated with population ageing in developed countries fall on these types of government expenditures on pensions, health and social services (Bloom et al., 2015; Rodrigues et al., 2018; Lyons et al., 2018).
In this study, we focus on pensions by considering the main determinants that are associated with early retirement decisions.3 Hence, some policies can help developing countries to encourage elder workers to stay longer in the labour market and so, delay retirement.4 The chances of policy success would definitely depend on a better understanding of ageing in the workforce and the particular role of health and work characteristics in those decisions (Van den Berg et al., 2010). We aim to contribute to this field of knowledge by studying the push and the pull factors influencing the retirement decisions across European countries. The research contribution of this paper is therefore the European view on the topic.
Previous studies have suggested the importance of health on early retirement decisions as an effort-reword strategy. That is, it is well known how illness is
1 Demographic aging, due to its structural nature, is a questionless process over the European Union.
According to Eurostat (2017), the proportion of people aged 65 or over rose from 23.5% to 29.3%
between 2001 and 2016 for the EU-28 countries.
2 The dependency ratio jointly with inactive population is creating new challenges regarding broader welfare policies. The number of retired people in Europe has increased significantly in recent decades and will continue to increase (De Preter et al., 2013).
3 The greatest urgency of the problem lies in the fact that the generations born in the Baby Boom years are getting the effective age of retirement.
4 According to results of the EU Labour force survey ad hoc module (2012) on the transition from work to retirement by Eurostat, on average in the EU-28, persons aged 50-69 who received an old-age pension and took part in an early retirement scheme were 58 years old when they retired. Nevertheless, labour market participation and retirement rates among older workers differ between European countries.
4
incompatible with working and so it is considered as a push factor. Hence, low health levels would be a reason that explains exits from the job market (Siegrist, 1996;
Karpansalo et al., 2004; Cai and Kalb, 2006). Besides, quality of work and working conditions has been also highlighted in recent times. Indeed, freshly contributions point out that the main factors that could explain early retirement are the ones associated with income or financial incentives, health conditions and quality of work (Siegrist et al., 2006). In this way, the lower the incentive, health status or quality of work, the higher the intention of retirement.
The main source of information used in this study is from the Survey on Health, Ageing and Retirement in Europe (SHARE) in order to contribute to the existing literature regarding early retirement decisions. Keeping all of that in mind, we have used all the existing waves (Waves 1 to 6: 2004-2015) in order to exploit the longitudinal dimension of the SHARE data. Therefore, using logistic regressions, we have tested for older employees the well-being and socioeconomic factors that determine the decision for retiring early from work. Our findings confirm previous contributions (Dendinger et al., 2005; De Preter et al., 2013; Carr et al., 2016; Moreira et al., 2018) and include factors associated with job satisfaction, quality of life and health. As a result, the major policy challenge would be to study those factors determining early retirement decisions in order to increase the number of employed people at older ages.5
The structure of the paper is as follows. In Section 2, we describe the data and methods based on the SHARE longitudinal survey. Besides, Section 3 shows the results whereas discussion, main conclusions and policy implications are presented in Section 4.
2. Data and Methods
Basic information is drawn from six waves (all the waves up to now) of the SHARE. It is a cross-national panel database of micro data customized to address multidisciplinary facets of ageing (health, socio-economic status and social and family networks) of more than 120,000 individuals of people age 50+ and over different European countries because it covers 27 European countries and Israel.
5 We do not make a distinction between “voluntary” and “involuntary” early retirement (for a profuse understanding see Dorn and Sousa-Poza (2010)).
5
Precisely, we take advantage of data from easySHARE release 6.0.0 (waves 1-6, since 2004/05 to 2014/15). However, our eligible sample is based on data availability.
Figure 1 describes the selection process showing that our sample is restricted to those employed persons aged 50-65 (n = 62,451). Next, we have excluded those data which does to lack of follow-up information understood as individuals who do not respond in consecutive waves (n = 56,558). Thus, our analytical sample consists of 5,893 individuals that are not for the full sample countries that comprise the SHARE project.
[Insert Figure 1]
Table 1 shows the sample distribution by country. The 9 countries that could be considered are thereabouts half of the total, and they represent Nordic (Sweden and Denmark), Continental (Austria, Germany, France, Switzerland and Belgium), and Southern Europe (Spain and Italy). Therefore, before reporting our results it is worth noting that, although SHARE database includes a rich set of information, there are not enough observations to disentangle by country. Hence, here we focus our attention on the completed SHARE panel that are based on 9 countries grouped in a sample size of 5,893 observations.
[Insert Table 1]
Summing up all these findings and in order to investigate the relationship between different factors and early retirement intentions among European workers aged 50-65, we have taken into account several sets of socio-economic determinants as exogenous variables following the results of previous studies (De Preter et al., 2013;
Carr et al., 2016) and their availability in the SHARE database. Thus, we have considered as dependent variable the intended retirement as a binary one that takes value 1 if the person is looking for early retirement in his/her job and zero otherwise.
Figure 2 plots the distribution of early retirement, considered as our dependent variable, by wave. Percentage values would decrease among waves, with a trend change in the latest one. That is, from 44.62% of early retirement intentions in wave 1 (2004/05) to 41.51% in wave 6 (2014/15) being the smallest one (38.93%) for wave 5 (2012/13).
[Insert Figure 2]
6
In order to investigate the push and the pull factors influencing the retirement decisions, we include different determinants (job characteristics, well-being and control factors) as exogenous variables following the results of previous contributions (Cantarero et al., 2017; Moreira et al., 2018) and their availability in the SHARE database: All variables used in estimates, which cover the whole relevant aspects, are described in Table 2 and calculated as follows.
[Insert Table 2]
Measures of working conditions: Information about it, is covered through two main variables: Job satisfaction and Civil servant. The first one takes value one when the worker is satisfied with his/her corresponding job (and zero otherwise), whereas the latest refers to the fact the employee is a civil servant. If that is the case, the variable takes value 1 and if not it takes value 0.
Measures of well-being: regarding it, there have been considered several set of factors. Quality of life (QoL) that is a commonly used measure for well-being, CASP- 12 in SHARE data, is usually stable across countries and time. It ranges between 12 and 48 and it is interpreted as follows: low QoL, <35; moderate, 35–37; high, 37–39; and very high, >= 39. Besides, some health factors attending several determinants like self- assessed health (SAH) by SAH-good or better and SAH-less than good, have limitations in daily activities (limited: 1 if the respondent reports any difficulties, 0 otherwise), body mass index (through overweight and obesity dummy variables), reporting any chronic condition (if respondent reports any chronic disease the variable chronic takes value 1, if not it takes value 0), and depression (if respondent has depression the variable would take value 1, 0 otherwise). As well, lifestyles are considered over if the respondent has ever smoked daily again using a dummy variable.
Socio demographic characteristics and control factors: we have considered age (three levels: 50-54 years; 55-59 years 60-65 years), gender (1 if female), education (measured according to the international classification ISCED-97: low, middle and high education understood as loweduc, mideduc and higheduc), geographic (rural: if the person lives in a rural area of location or not) and social isolation (alone: if the person lives alone or not) and we have used it as covariates.
7
Moreover, due to our dependent variable is a binary one, logistic regression models are used to analyse the impact of job characteristics, well-being, and control factors on early retirement decisions among the youngest-old people for our sample of European countries using SHARE (Jones et al., 2013; Deb et al., 2017).6 We consider the Model I that include baseline variables as job satisfaction, civil servant, gender, age and education and the Model II that includes all variables (Model I plus CASP, SAH, overweight, obesity, chronic, depression, eversmoked, rural and alone). The statistical analysis using the full sample (9 countries) is performed using Stata14 (Rabe-Hesketh and Skrondal, 2008).
3. Results
We begin with a simple explanation about the sample of participants that with full data consists in 5,893 individuals being 51.67% females. The age distribution is as follows:
26.57% are <55 years old; 40.95% are 55-59 years old, and 32.48% are 60-65 years old.
Descriptive statistics for the analytical sample are showed in Table 3. We can observe that prevalence among participant shows huge differences when we are looking for early intended retirement decisions. Besides, as we expected, satisfied with main job, very high CASP, good of better SAH or individuals living alone would have less intentions of early retirement. This point can best be appreciated by noting that this behaviour is more intense for males, people aged 60-65 years old near to legal retirement and highly educated. Hence, Table 3 is the first approximation to determine the push and pull factors associated with early retirement decisions.
[Insert Table 3]
Next, Table 4 shows the results for logistic models where odds ratios of intended early retirement are used. This statistic is commonly used in the literature to present the results of health analysis as an alternative way to express possibility for the occurrence of an outcome or presence of an exposition. That is, for each variable, the likelihood is compared with the reference group.
6 Logit (logistic regression) is more popular in health sciences, like epidemiology, as coefficients can be interpreted in terms of odds ratios. Here those results are presented in the following section.
8
[Insert Table 4]
Regarding the first OR results column, our findings are as expected and advanced in Table 3. Statistically significant results are obtained for job satisfaction, civil servant status, age and educational variables. Nonetheless, the gender variable (being female) is not significant.7 Therefore, 1.29 (odds ratio) means that the odds of retiring early represent a 29% higher for civil servant responders. However, smaller odds are obtained for being satisfied with (main) job, the eldest age cohort considered and the most educated individuals. That responders were a half as likely to retire early as those non-satisfied with their job, younger individuals or less educated people (95%
C.I.: 0.48 to 0.85).
Furthermore, if we consider all the variables simultaneously (second OR results column), results are just about unchanging with the aforementioned ones. Elder workers’ retirement odds increases with age and low well-being. For example, regarding SAH, if it is good or better (OR = 0.55; 95 % C.I. 0.47, 0.65) being about 50% more likelihood of staying at work. This possibility is also higher for civil servant employees around 30% more chance of staying at work and not leaving from paid employment (OR = 1.29; 95 % C.I. 1.12, 1.48). Next, women differ from men slightly and tend to retire earlier than they do (OR = 1.12; 95 % C.I. 0.99, 1.25). The respondent’s with higher education has important significant effects, which means that older workers with a higher educational level are likely to remain on the labour market longer than lower-educated workers (OR = 0.55; 95 % C.I. 0.48, 0.64). Rurality, limitations, chronic or depression factor are not statistically significant.8
From the presentation above, it could be argue that a person satisfied with his/her job, not being civil servant and male, with high education or quality of life, with good or better SAH, with no weight related problems, that has never smoke and lives alone, is less likely to be a worker that is looking for early retirement in their main job.
7 Previous empirical literature has shown that Evidence that couples coordinate the timing of retirement (Lee, 2017).
8 However, interventions encouraging work participation of elderly workers should take into account possible differences between groups (i.e. with and without chronic diseases (Sewdas et al., 2018)),
9
4. Discussion and Conclusions
Given the challenge of aging population on welfare states, over the last years many studies were focused on issues related to older employees who retire early from work.
That is, following recent contributions on this area (De Preter et al., 2013; De Wind et al., 2015; Carr et al., 2016; De Wind et al., 2017), we have determined which are the push and the pull factors influencing the retirement decision of older workers (age 50+) are.
We have analysed associations of indicators of well-being and the decision on early retirement at the individual level. In doing so, we have used data from the SHARE that collects information on both the economic and non-economic characteristics of the youngest-old European people. Indeed, we have worked with a longitudinal sample of older male and female employees in nine European countries.
Overall, our empirical findings indicate that socioeconomic and well-being factors would affect the decision on premature departure from work. Starting with the pull factors, here it is highlighted that job satisfaction clearly matters, around 60% more chance of staying at work. This result was previously found in studies like the ones of Dendinger et al. (2005) or Carr et al. (2016). Similar results are shown, as in Siegrist et al. (2006), for educational variables and we have considered those socioeconomic factors which are behind it; or in case of living alone, we have hypothesised that this fact would be a reason of non-familiar responsibilities.
Besides, the fact being female and/or a civil servant employee would proceed as push factors. At this regard, Mein et al. (2000) found that the longer a person (both men and women) works in the civil service the more likely they are to retire early. Our results are similar to previous contributions that have tested how health is the most commonly push factor for early retirement. Accordingly, workers who are eligible for early retirement may opt for exists the job market rather than further exposure to adverse working conditions associated with health deteriorations (Jones et al., 2013).
Our findingswhile regarding issues associated with SAH, quality of life or body mass index related factors show statistically significant effects (the same applies for our lifestyle variable: ever smoke), for limitations, chronic or depression there are not statistically significant results. In addition, there are not so many differences between respondents living in a small town, a rural area or village and people from the city.
10
Therefore, our empirical analysis is concerned with the individual level and we can conclude that there are push and pull factors influencing work-retirement transitions of elderly workers (De Preter et al., 2013). Some limitations and extensions should be indicated. Firstly, in spite of considering individual-specific characteristics, we should not forget that we are working with self-reported information. This fact could have implications for results and consequently, policy recommendations should be taken with caution. Nevertheless, possible biases in self-evaluation are partly explained by using more controlled variables as it has been done in this study. Secondly, upcoming studies require explore both the evolution on individual level and differences by countries in order to gain a better understanding for coordinated social policies. Here, we do not have enough observations to disentangle by country. Thus, when more data would be available, we could focus on that issue and consider clusters which visualize very intuitively the relative weights given by each country.
Despite the limitations of this analysis, we can confirm that the longitudinal SHARE data used here provide new information on the impact of well-being on intended early retirement, as well as on the different ways in which it affects people according to their job, gender, age, health, or social-environment. All these issues constitute an essential tool for policymakers when designing policies that target pension systems among European countries. That is, along with traditional sustainability policies regarding pension systems such as increasing the effective age of retirement, alternatives associated with pull factors as work environment factors or flexibility, among others should be considered in the near future. Without doubt, the sucess of these strategies would depend on a better understanding of ageing in the workforce and the importance of the above-mentioned factors.
References
Bloom, D. E., Chatterji, S., Kowal, P., Lloyd-Sherlock, P., McKee, M., Rechel, B., ... &
Smith, J. P. (2015). Macroeconomic implications of population ageing and selected policy responses. The Lancet, 385(9968), 649-657.
Börsch-Supan, A., S. Gruber, C. Hunkler, S. Stuck, S. Neumann, J. (2017).
EasySHARE. Release version: 6.0.0. SHARE-ERIC. Dataset. doi:
10.6103/SHARE.easy.600
11
Cai, L., & Kalb, G. (2006). Health status and labour force participation: evidence from Australia. Health Economics, 15(3), 241-261.
Cantarero, D., Pascual, M., Blazquez, C. (2017). What is Happening with Quality of Life Among the Oldest People in Southern European Countries? An Empirical Approach Based on the SHARE Data. Social Indicators Research, in press.
Carr, E., Hagger-Johnson, G., Head, J., Shelton, N., Stafford, M., Stansfeld, S., &
Zaninotto, P. (2016). Working conditions as predictors of retirement intentions and exit from paid employment: a 10-year follow-up of the English Longitudinal Study of Ageing. European Journal of Ageing, 13(1), 39-48.
Deb, P., Norton, E.C., Manning, W.G. (2017). Health econometrics using STATA. Stata Press.
Dendinger, V. M., Adams, G. A., & Jacobson, J. D. (2005). Reasons for working and their relationship to retirement attitudes, job satisfaction and occupational self- efficacy of bridge employees. The International Journal of Aging and Human Development, 61(1), 21-35.
De Preter, H., Van Looy, D., & Mortelmans, D. (2013). Individual and institutional push and pull factors as predictors of retirement timing in Europe: A multilevel analysis. Journal of Aging Studies, 27(4), 299-307.
De Wind, A., Geuskens, G. A., Ybema, J. F., Bongers, P. M., & van der Beek, A. J.
(2015). The role of ability, motivation, and opportunity to work in the transition from work to early retirement–testing and optimizing the Early Retirement Model. Scandinavian Journal of Work, Environment & Health, 24-35.
De Wind, A. D., Burr, H., Pohrt, A., Hasselhorn, H. M., Van der Beek, A. J., &
Rugulies, R. (2017). The association of health and voluntary early retirement pension and the modifying effect of quality of supervision: Results from a Danish register-based follow-up study. Scandinavian Journal of Public Health, 45(5), 468-475.
Dorn, D., & Sousa-Poza, A. (2010). ‘Voluntary’and ‘involuntary’early retirement: an international analysis. Applied Economics, 42(4), 427-438.
Eurostat (2012). Ad hoc modules of the European Union labour force survey (EU-LFS):
Transition from work into retirement. Brussels, Eurostat.
Eurostat (2017). Eurostat Statistics explained. European Commission.
12
Gruber, S., Hunkler, C., Stuck, S. (2014). Generating easySHARE Guidelines, Structure, Content and Programming. Working Paper Series 17-2014, SHARE Working Paper Series.
Jones, A. M., Rice, N., d'Uva, T.B. & Balia, S. (2013). Applied health economics. 2nd Edition, Routledge.
Jones, M. K., Latreille, P. L., Sloane, P. J., & Staneva, A.V. (2013). Work-related health risks in Europe: are older workers more vulnerable? Social Science & Medicine, 88, 18-29.
Karpansalo, M., Manninen, P., Kauhanen, J., Lakka, T. A., & Salonen, J.T. (2004).
Perceived health as a predictor of early retirement. Scandinavian Journal of Work, Environment & Health, 287-292.
Lee, A. (2017). Late Career Job Loss and Retirement Behavior of Couples. Research on Aging, 39(1), 7-28.
Lyons, A.C., Grableb, J.E., Jooc, S. (2018). A cross-country analysis of population aging and financial security. The Journal of the Economics of Ageing, 12, 96- 117.
Mein, G., Martikainen, P., Stansfeld, S. A., Brunner, E. J., Fuhrer, R., & Marmot, M.G.
(2000). Predictors of early retirement in British civil servants. Age and Ageing, 29(6), 529-536.
Moreira, A., Azevedo, A. B., & Manso, L.P. (2018). Reducing early retirement in Europe: do working conditions matter? Journal of Population Ageing,11(3), 265-284.
Rabe-Hesketh, S., & Skrondal, A. (2008). Multilevel and longitudinal modeling using Stata. STATA press.
Rodrigues, R., Ilinca, S., & Schmidt, A.E. (2018). Income‐rich and wealth‐poor? The impact of measures of socio‐economic status in the analysis of the distribution of long‐term care use among older people. Health Economics, 27(3), 637-646.
Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal of Occupational Health Psychology, 1(1), 27-41.
Siegrist, J., Wahrendorf, M., Von dem Knesebeck, O., Jürges, H., & Börsch-Supan, A.
(2007). Quality of work, well-being, and intended early retirement of older employees—baseline results from the SHARE Study. The European Journal of Public Health, 17(1), 62-68.
13
Sewdas, R., van der Beek, A. J., de Wind, A., van der Zwaan, L. G., & Boot, C. R.
(2018). Determinants of working until retirement compared to a transition to early retirement among older workers with and without chronic diseases: Results from a Dutch prospective cohort study. Scandinavian Journal of Public Health, 46(3), 400-408.
Van den Berg, T., Schuring, M., Avendano, M., Mackenbach, J., & Burdorf, A. (2010).
The impact of ill health on exit from paid employment in Europe among older workers. Occupational and Environmental Medicine, 67(12), 845-852.
14
TABLES AND FIGURES
Figure 1 Flow chart of the analytical sample
Source: Authors’ elaboration based on easySHARE release 6.0.0 (Waves 1 to 6: 2004-2015).
Notes: Current job situation is not available from Wave 3.
Aged 50-65 years old (n = 135,916)
Eligible sample (n = 62,451)
Analytical sample (n = 5,893 ) Lost do to lack of
follow up (n = 56,558) Excluded because current work situation was not employed or self-employed
(n = 73,462)
15
Table 1 Distribution of the analytical SHARE sample by country (all countries (9);
sample size (n) = 5,893)
Country/Wave (years)
2004/05 2006/07 2010/2011 2012/13 2014/15 Total
Austria 68 64 38 20 10 200
Belgium 340 309 170 131 81 1,031
Denmark 272 249 195 147 103 966
France 222 195 108 74 46 645
Germany 195 164 112 81 57 609
Italy 169 129 78 52 36 464
Spain 131 107 78 50 32 398
Sweden 344 306 182 126 79 1,037
Switzerland 164 149 103 82 45 543
Total 1,905 1,672 1,064 763 489 5,893
Source: Authors’ calculations based on easySHARE release 6.0.0 (Waves 1 to 6: 2004-2015). Population of workers aged 50-65.
Notes: Current job situation is not available from Wave 3.
16
Figure 2 Distribution (percentages) of early retirement by SHARE Wave
Source: Authors’ calculations based on easySHARE release 6.0.0 (Waves 1 to 6: 2004-2015). Population of workers aged 50-65.
Notes: Current job situation is not available from Wave 3.
55.38
44.62
61.18
38.82
60.06
39.94
61.07
38.93
58.49
41.51
02040600204060
-1 0 1 2
-1 0 1 2 -1 0 1 2
2004-2005 2006-2007 2010-2011
2012-2013 2014-2015
Percent
Looking for early retirement in (main) job
Graphs by Wave
17
Table 2 List of variables and description
Variable Description Coding
Dependent variable Early retirement Worker is looking for early retirement in (main) job 1: yes; 0: otherwise
working conditions Job satisfaction Worker is satisfied with (main) job 1: agree or strogly agree; 0: otherwise Civil servant Worker is a civil servant 1: civil servant; 0: otherwise
well-being
CASP Quality of life (QoL)
The CASP-12v Quality of life and well-being index. Each of its 12 items are answered using a four-point Likert-type scale, and the total score, which ranges between 12 and 48 is interpreted as follows: low QoL, <35; moderate, 35–37; high, 37–39; and very high, >= 39.
SAH-good or better Self-perceived health, good or better 1: good or better; 0: otherwise SAH-less than good Self-perceived health, less than good 1: less than good; 0: otherwise
Limited Activities of daily living 1: respondent reporting any difficulties; 0: otherwise Overweight Body Mass Index: 25-29.9 in kg/m2 1: respondent is overweight; 0: otherwise
Obesity Body Mass Index: 30 or above in kg/m2 1: respondent is obese; 0: otherwise
Chronic Chronic diseases 1: respondent reporting any chronic disease; 0: otherwise
Depression Depression 1: respondent has depression; 0: otherwise
Eversmoked Whether respondent has ever smoked daily 1: respondent has ever smoked daily; 0: otherwise
socio-demographic characteristics and control factors
Female Gender of respondent 1: female; 0: male
Age Age of respondent Years
Loweduc ISCED-97 coding of education, low education 1: low education; 0: otherwise Mideduc ISCED-97 coding of education, middle education 1: middle education; 0: otherwise Higheduc ISCED-97 coding of education, high education 1: high education; 0: otherwise
Rural Area of location (place of residence) 1: respondent lives in a small town, a rural area or village; 0:
otherwise
Alone Number of people living in the respondents’ household 1: respondent live alone; 0: otherwise Source: Authors’ elaboration.
18
Table 3 Descriptive statistics of the analytical sample (all countries (9); sample size (n)
= 5,893)
Variable Total Early intended
retirement (%) Mean S.D. (%) Mean S.D.
working conditions
Job satisfaction 83.08 0.83 0.37 79.17 0.79 0.41 Civil servant 19.36 0.19 0.40 20.79 0.21 0.41
well-being
CASP-low 24.15 0.24 0.43 31.19 0.31 0.46 CASP-moderate 7.42 0.07 0.26 9.12 0.09 0.29 CASP-high 9.89 0.10 0.30 11.22 0.11 0.32 CASP-very high 58.54 0.59 0.49 48.47 0.48 0.50 SAH-less than good 12.86 0.87 0.33 18.85 0.81 0.39 SAH-good or better 87.14 0.13 0.33 81.15 0.19 0.39 Limited 2.36 0.02 0.15 3.01 0.03 0.17 Overweight 40.73 0.41 0.49 42.33 0.42 0.49 Obesity 16.44 0.16 0.37 19.55 0.20 0.40 Chronic 42.90 0.43 0.49 45.59 0.46 0.50 Depression 33.02 0.33 0.47 36.72 0.37 0.48 Eversmoked 51.40 0.51 0.50 54.00 0.54 0.50
socio- demographic characteristics
and control factors
Female 51.67 0.52 0.50 53.05 0.53 0.50 50-54 years 26.57 0.27 0.44 29.41 0.29 0.46 55-59 years 40.95 0.41 0.49 43.94 0.44 0.50 60-65 years 32.48 0.32 0.47 26.65 0.27 0.44 Loweduc 26.51 0.27 0.44 32.34 0.32 0.47 Mideduc 36.72 0.37 0.48 38.37 0.38 0.49 Higheduc 36.13 0.36 0.48 28.80 0.29 0.45 Rural 55.88 0.56 0.50 57.71 0.58 0.49 Alone 13.00 0.13 0.34 11.59 0.12 0.32
Source: Authors’ calculations based on easySHARE release 6.0.0 (Waves 1 to 6: 2004-2015). Population of workers aged 50-65.
19
Table 4
Associations of job satisfaction, socio-economic variables, health and early retirement intentions: logistic regressions models (odds ratios and 95% confidence intervals) for all countries (n = 5,893)
Model I Model II
Independent variables OR 95%CI OR 95%CI
working conditions
Job satisfaction
Yes 0.58 [0.50-0.61] *** 0.61 [0.53-0.71] ***
No 1.00 1.00
Civil servant
Yes 1.29 [1.12-1.48] *** 1.29 [1.12-1.48] ***
No 1.00 1.00
well-being
CASP
CASP-low 1.00
CASP-moderate 0.93 [0.74-1.16]
CASP-high 0.87 [0.72-1.07]
CASP-very high 0.56 [0.49-0.64] ***
SAH
SAH-less than good 1.00
SAH-good or better 0.55 [0.47-0.65] ***
Limited
Yes 1.03 [0.72-1.48]
No 1.00
Overweight
Yes 1.25 [1.10-1.41] ***
No 1.00
Obesity
Yes 1.39 [1.18-1.63] ***
No 1.00
Chronic
Yes 1.03 [0.92-1.15]
No 1.00
Depression
Yes 1.10 [0.98-1.24]
No 1.00
Eversmoked
Yes 1.14 [1.02-1.28] **
No 1.00
20 socio-demographic
characteristics and control factors
Gender
Female 1.07 [0.96-1.19] 1.12 [0.99-1.25] *
Male 1.00 1.00
Age
50-54 years 1.00 1.00
55-59 years 0.90 [0.79-1.03] 0.92 [0.81-1.05]
60-65 years 0.55 [0.48-0.64] *** 0.60 [0.52-0.70] ***
Education
Loweduc 1.00 1.00
Mideduc 0.74 [0.65-0.85] *** 0.82 [0.72-0.94] ***
Higheduc 0.47 [0.40-0.54] *** 0.55 [0.48-0.64] ***
Rural
Yes 1.03 [0.92-1.15] 1.01 [0.90-1.13]
No 1.00 1.00
Alone
Yes 0.82 [0.70-0.97] ** 0.79 [0.67-0.93] ***
No 1.00 1.00
McKelvey & Zavoina’s R2
0.05 0.10
Notes: ***,** and * indicate significance at 1%, 5% and 10%, respectively.
